Re: [webrtc-stats] Stat for audio playout delay

This is related to the need that was the basis for
The critical part of a definition here are the measurement points: At 
what point do we start the clock running, and at what point do we say 
"the delay is completed"?

Side point: If we can't get exact values (such as delays in the audio 
subsystem after the samples have been handed over), should we add an 
estimate into the delay, or use the last time at which we were able to
 measure time?

GitHub Notification of comment by alvestrand
Please view or discuss this issue at 
using your GitHub account

Received on Tuesday, 7 February 2017 15:06:04 UTC